Social Structure and Change in Public Opinion: the Case of Same-Sex Marriage

Over the past several decades, there has been a remarkable change in public policy about same-sex marriage in the United States, reflective of broader positive trends in attitudes toward lesbian, gay, and bisexual (LGB) people. In 1996, the US Congress overwhelmingly passed the Defense of Marriage Act (DOMA), which defined marriage as the “union of one man and one woman;” less than twenty years later, the US Supreme Court decided, in Obergefell v. Hodges (2015), that all states must issue marriage licenses to same sex couples.  Behind this policy change has been an extraordinarily rapid transformation in public attitudes about same-sex marriage:  according to Pew Research data, in 2001 57% of Americans opposed allowing gays and lesbians to marry legally whereas only 37% favored same-sex marriage; by 2015, these numbers had nearly reversed, with 55% in favor and only 39% opposed.

Though political entrepreneurship and legal strategizing no doubt played an important role in the success of the marriage equality movement, many commentators identify rapid change in the public's acceptance of gay marriage as a critical factor in the legislative and judicial victories that preceded Obergefell.  And yet across a wide range of other value-related issues, public opinion is slow to change.  Thus we must ask, what caused the dramatic change in American's attitudes toward same-sex marriage?

In this talk I will describe an empirically grounded, simulation-based investigation into the rapid change in public opinion about same-sex marriage in the US.  We focus on two factors that differentiate the same-sex marriage issue from other civil rights movements: the practice of closeted gay Americans “coming out,” (and thereby suddenly making a hidden trait visible to those in their network), and the essentially random distribution of gay people within families (but not networks).  We show that these two factors are capable of producing a cascade of opinion change, a pattern that we contrast with the slower rise in support if the most mobilized populations are either visible or clustered together in networks.

Our paper makes two contributions.  First, it explores the limits of a theoretically informed mechanism that plausibly accounts for an important recent social change in the US:  the rise in support of marriage equality.  Second, it serves as a powerful illustration of how micro-processes operate on networks, and how network structures enable or constrain macro-level changes.